For what is intended (automobile powered ozone generation) either will work. The efficiency difference is under a watt.
The trick is providing a steady stream of relatively dry air.
I work with bio-medical ozone devices and we simply run an adsorption oxygen generator followed with a condensate scrubber and topped off with a polymer absorption canister like those beads you find in some food packets to absorb moisture.
You don't need a huge amount of ozone. Simply calculate your air consumption at cruise and translate into molar mass. Then, you just need an equivalent of 40 parts per million to greatly accelerate combustion processes. Even concentrations in parts per billion have measurable effects.